DESCRIPTION (Investigator's Abstract): Type 2 diabetes is one of the most common metabolic disorders in humans and has a complex etiology due to environmental and genetic factors. Recently, a positional candidate region, NIDDM1, has been proposed to contain variation which contributes to diabetes risk. According to the thrifty genotype hypothesis, diabetes susceptibility genotypes conferred a selective advantage in the ancient past in times of feast a famines. In this application, we propose to conduct a detailed study on the population and evolutionary genetics of the candidate region for the genetic susceptibility to type 2 diabetes. The ultimate goals of our study are to provide information on the haplotype structure of variation that can be used to design and interpret replication studies, understand the forces that shape sequence variation and LD at this locus in aboriginal human populations, and use the signature of natural selection on this region to validate the mapping evidence. To advance these goals, we propose to: Survey sequence variation in small random samples (10 individuals) from each of four major ethnic groups (Africans, Asians, Europeans, and Native Americans) in the two genes found in the NIDDM1 candidate susceptibility region. The data generated in this survey will be analyzed to identify sub-regions with evidence for positive natural selection. These regions in addition to those containing candidate diabetes susceptibility variants will be further studied in Specific Aim 2. Conduct a detailed analysis of sequence variation and LD in a sample of 25 individuals each from large outbred populations from three major ethnic groups (Africa, Asia, and Europe). We will use these data to provide critical information for disease mapping and look for the signature of natural selection based on the "thrifty" genotype hypothesis. Carry out an extensive survey of allele frequencies at NIDDM1 candidate susceptibility variants in 20 human population samples with a broad range of ethnic and geographic origins and rates of type 2 diabetes prevalence. The survey results will be analyzed to ask if the at-risk genotype frequencies correlate with diabetes prevalence rates In different populations. In addition, we will be able to determine whether the degree of inter-population differentiation is significantly different from that observed at a large number of neutrally evolving nuclear loci and whether the geographic distribution of allele frequencies is correlated with environmental features, such as latitude or climate.